This is a suite which enables to calculate the enrichment ratios of 100,000 permuted sets of CoRSIVs and control probes.
-
Data- This has all the input files for the permutation testing including the probe lists for each disease. -
permuted_sets- This includes 100,000 files. Each files contains permuted CoRSIV and control probes. This is the output ofget_permutations.py. Due to the large size of this file, it was not added here. It's located in the SPHERE cluster (/home/maduranga/Permutation_testing_code_SPHERE). -
p1.pytop10.py- Calculate enrichment ratios for 100,000 permuted sets. Each file processes 10,000 sets which can be run in parallel if needed. -
merge_chunks.py- Concatenates the chunks to create one final output for each disease.
- Add the probe list CSV file of the disease to
Datafolder.
- The file name should end with "_all_probes.csv". Ex:
neurological_all_probes.csvorInsulin Resistance_all_probes.csvorDiabetes Mellitus, Type 2_all_probes.csv
- Use the
permutation_testing.slrumto run permutation testing.
- Example:
#!/bin/sh
#SBATCH --time=48:00:00 -n24 -p bynode
for d in "neurological"
do
python p1.py "${d}"
python p2.py "${d}"
python p3.py "${d}"
python p4.py "${d}"
python p5.py "${d}"
python p6.py "${d}"
python p7.py "${d}"
python p8.py "${d}"
python p9.py "${d}"
python p10.py "${d}"
done
You can run each of p1.py to p10.py in parallel.
- Example:
#!/bin/sh
#SBATCH --time=24:00:00 -n24 -p bynode
python p1.py "neurological"
#!/bin/sh
#SBATCH --time=24:00:00 -n24 -p bynode
python p2.py "neurological"
#!/bin/sh
#SBATCH --time=24:00:00 -n24 -p bynode
python p10.py "neurological"
- After the above script finishes running, merge the results by running
merge_chunks.slrum.
- Example:
#!/bin/sh
#SBATCH --time=48:00:00 -n24 -p bynode
python merge_chuncks.py "neurological"
The output will be generated in final_permutation_testing_results folder.